545 Deep learning/artificial intelligence for automatic measurement of global longitudinal strain by echocardiography

Funding Acknowledgements

The Norwegian Health Association, South-Eastern Norway regional health Authority and the national program for clinical therapy research (KLINBEFORSK).

Background

Global longitudinal strain (GLS) by echocardiography has incremental prognostic value in patients with acute myocardial infarction and heart failure compared to left ventricular (LV) ejection fraction and provides more reproducible measurements of LV function. Recent advances in machine learning for image analysis now open the possibility for robust fully automated tracing of the LV and measurement of global longitudinal strain (GLS), without any operator input. This could make real-time GLS possible and remove inter-reader variability, thus resulting in saved time and improved test-retest reliability. The aim of the present study was to investigate how measurements by this novel automatic method compares to conventional speckle tracking analyses of GLS.

Methods

100 transthoracic echocardiographic examinations were included from a clinical database of patients with acute myocardial infarction or de-novo heart failure. Examinations were included consecutively and regardless of image quality. Simpson biplane LV ejection fraction ranged from 7 to 70%. Images of three standard apical planes from each examination were analysed using our novel and fully automated GLS method based on deep learning technology. The automated GLS measurements were compared to conventional speckle tracking GLS measurements of the same acquisitions using vendor specific format and software (EchoPAC, GE Healthcare), performed by a single experienced observer.

Results

GLS was -11.6 ± 4.5% and -12.8 ± 5.0% for the deep learning method and the conventional method, respectively. Bland-Altman analysis showed a bias of -0.7 ± 1,9% and 95% limits of agreement of -4,6 to 3.1. No clear value dependent bias was found by visual inspection (Figure A). Feasibility for measurement of GLS was 93% for the deep learning based method and 99% for the conventional method. The limits of agreement found in our study is comparable to findings in the intervendor comparison study by the EASCVI/ASE/Industry Task force to standardize deformation imaging.

Conclusion

This novel deep learning based method succeeds without any operator input to automatically identify and classify the three apical standard views, trace the myocardium, perform motion estimation and measure global longitudinal strain. This could further facilitate the clinical use of GLS as an important tool for enhancing clinical decision-making.

Abstract 545 Figure.

P1428 Mortality in non-ischemic cardiomyopathy is low and close to the general population

Funding Acknowledgements

South Eastern Norway regional health authority

Background

Patients with non-ischemic dilated cardiomyopathy (NDCM) have lower mortality compared to patients with ischemic cardiomyopathy (ICM). Recent reports suggest less benefit of a primary prophylactic implantable cardioverter defibrillator (ICD) in NDCM.

Purpose

We aimed to investigate mortality rate and appropriate ICD therapy in a consecutive cohort of patients with NDCM and ICM.

Methods

In this prospective multi-center study, we consecutively included NDCM and ICM patients with left ventricular ejection fraction (LVEF) < 40% from July 2014 to January 2018. ICM or NDCM classification was based on coronary angiography. Echocardiography was performed at inclusion and LVEF and global longitudinal strain (GLS) were assessed. All-cause mortality and primary prevention ICD therapies were recorded during follow-up.

Results

We included 290 patients (67 ± 13 years old, 74% males), 207 with ICM and 83 with NDCM. At inclusion LVEF was 31 ± 6% and GLS -10.5 ± -3.3%. Patients with ICM were older (68 ± 12 years vs. 63 ± 15 years, p < 0.01), had better LVEF (32 ± 6% vs. 28 ± 7%, p < 0.01), and shorter QRS duration (106 ± 23 ms vs. 122 ± 28 ms, p < 0.01). A primary prevention ICD was implanted in 18 (9%) ICM patients and in 21 (25%) NDCM patients (p < 0.001).

During 22 ±12 months follow up, all-cause mortality was 27 (9%) in the entire population. In patients with ICM 1/18 (6%) received appropriate shock from their primary prevention ICD compared to 3/21 (14%) NDCM patients. Mortality was more frequent in ICM; 26/207 (13%) compared to 1/83 (1.2%) in NDCM (Log rank p < 0.01). All-cause mortality or appropriate ICD shock was more frequent in ICM compared to NDCM (27 (13%) vs. 4 (5%) log rank p = 0.02) (Figure). Assuming the arrhythmias treated by the ICD shocks would have been lethal, annual mortality of 2.5% in the NDCM population was only slightly higher compared to the age matched general population (1% annual mortality) (blue line).

Conclusion

Patients with NDCM had better survival, and in general a lower event rate, compared to patients with ICM. However, subgroups of NDCM patients may benefit from their primary prophylactic ICD and further studies should investigate the need of primary prevention ICD in patients with NDCM.

Abstract P1428 Figure.